G protein-coupled receptors (GPCRs) are integral membrane proteins that transduce extracellular signals across the cell membrane and have been directly linked to several diseases including cancer, diabetes and HIV infection. GPCRs comprise the largest class of membrane proteins in mammalian cells - the human genome encodes more than 800 - and recent estimates indicate that GPCRs represent 60-70% of all drug targets. Yet despite their widespread occurrence and potential medical relevance, the structure, function and biology of most GPCRs remains elusive due to difficulties in protein expression, purification and crystallization, as well as the general lack of reagents for their manipulation in native environments. As a result, only four unique GPCR structures have been reported. In several of these cases, GPCR structure determination was greatly aided by specifically binding antibodies and antibody fragments (e.g., Fab or Fv). Unfortunately, antibodies that bind specifically and tightly to integral membrane proteins have been challenging to generate using conventional means. Hence, the goal of this project is to develop new genetic and protein engineering tools that allow fast access to high-affinity antibodies that bind GPCRs. Specifically, two novel protein-fragment complementation assays (PCAs) will be engineered in E. coli for detecting protein-protein interactions involving GPCRs. These PCA strategies are based on the principle of split protein complementation, in which two inactive fragments derived from an enzyme are fused to a pair of interacting partners. When the two partners interact, the two inactive fragments are brought into proximity and an intact enzyme is reconstituted. The first strategy employs TEM-1 2-lactamase PCA to detect interactions between periplasmically expressed antibody fragments and native epitopes of GPCRs expressed in the cytoplasmic membrane (specific aim 1). The second strategy uses yeast cytosine deaminase PCA to report antibody fragment-mediated blocking of GPCR dimerization (specific aim 2). In each of these selection strategies, the coding sequence of the GPCR of interest is fused to a protein fragment derived from a selectable marker enzyme. Consequently, the reassembly of the marker and survival on selective media is dependent upon an interaction between the GPCR and a partner protein fused to the other half of the selectable marker. The advantage of this approach is that molecular interactions involving GPCRs can be directly and rapidly detected in the context of living E. coli cells using simple clonal selection. Studies here will focus on the following mammalian GPCRs from the rhodopsin family of receptors: adenosine A2A, cannabinoid CB1 and CB2, the long splice variant of dopamine D2 and neurotensin receptor-1. The development of convenient genetic tools for rapidly isolating GPCR- specific antibodies in a high-throughput and cost-effective way would address a major technological gap that has hampered the GPCR field and the NIH scientific community as a whole.